Need experts for vexing hum problem

There is no hum when the preamp is switched off (butt connected and plugged in.) I replaced the original switch with the little neon lamp. The switch is now in the remote box. There is a pic of it opened up here:

The black rectangle is the back of a 3A fuse holder.

As far as connections, I hope I understand you here, I get the hum with the inputs shorted or with an ipod connected (but not plugged in, running on battery power.) I do not have my probes connected or any other gear plugged in. The hum is always consistent in volume and equal in both channels regardless of volume adjustment. (I have removed the balance pot.) Putting the lid on does not effect the hum at all. No scope or DMM connected. With the amp already on the hum is at full volume as soon as I throw the preamp switch. And again, with or without tubes and/or rectifier tube.

thanks for helping,

gary

So this is yet another recommendation to accommodate an earth ground to my amp. I'll get on it presently. I have an Adcom GFA 5500, is anywhere inside on the chassis suitable for the earth connection?

thanks,

gary

Paul,

rereading your last post I see you mean that I should install 3-prong outlets in my apartment. Believe me, I would if I could. As it is I have run an extension from the grounded outlet in the kitchen, which, again, warrants converting my amp to have an earth ground.

gary

I appreciate that safety is of some concern. This is particularly true of higher voltage circuits, with vacuum tubes in 'em and marketed by scoundrels, grounded or not. On a test bench, a low impedance ground can be more of a hazard than a safety feature ( to the technician or his stray relatives) if not carefully identified and deployed.

The OP says, in this thread, that he's tapped a ground wire off the stove for third wire uses where deemed suitable; that he's using an isolation transformer for the stereo.

Let's fix the amplifier first. Rewiring the house probably ought to wait for better weather.

RL

I must have missed that one. Just the same the preamp has a three prong plug. Go figure.

Without absolutely accurate schematics it is difficult to do much better than guess.

Just the same there are serious issues with the PA design.

Replace your Ipod on the preamp inputs with shorting plugs until hum is solved. All other unused inputs on the preamp and power amp should also be shorted.

Where is the previously mentioned isolation transformer located? Only the preamp and power amp should be plugged into it, with no ground wire from the isolation transformer casing or third wire socket pin connected. Bring out this ground wire/isolation transformer case connection to a lead wire for temporary attachment to hardware. Leave the power amp on.

At this time:

1) Note effect on hum (sw off / sw on - in each case ) of a ground wire connection to 0 1 a) b) - the power amp chassis only c) d) - the preamp chassis only.

If there is a reduction in hum for any of the four conditions, note it down.

2) Disconnect the preamp outputs from the power amp inputs and plug your Ipod directly into both inputs of the power amp. ( A realistic power amp input volume setting should be established that will be maintained until hum is defeated.) There should be no hum at this volume level. At maximum levels, without an program signal, you are likely to hear breathing, hum, clicks and possibly SW radio.

a) b) Leave the ipod connected to the power amp inputs. Switch on/off the disconnected preamp. There should be no effect.

3) Leave the Ipod connected to the power amp inputs. With the disconnected preamp on, brush the outer shell (signal return) on the end of one of the preamp's output audio connectors against the power amp chassis. Note the effect.

4) Repeat this with the ground wire attached to

a) - the power amp chassis only. b) - the preamp chassis only.

5) Leave the ipod connected to the power amp left input channel only. With the disconnected preamp on, brush the outer shell of the power amp right channel input patch cord against the preamp chassis.

6) Repeat this with the ground wire attached to

a) - the power amp chassis only. b) - the preamp chassis only.

Share the results and we'll all think about it for a while.

RL

I still suspect the DC power supply for the filamants. The capacitors all connect to the ground bus on the preamp PCB, along with the signal grounds. This circuit should be isolated, or at most connected at one place, such as the transformer CT. It is important to keep power grounds and signal grounds separated. Capacitor charging currents can have peaks of 100 amps or more, and even a 1 milliohm resistance in the ground circuit can create

100 mV of AC noise.

One clue that points to this possibility is that you get the hum with all the tubes removed, so this circuit is the only one that is still operating. Disconnect one of these transformer wires and see if the problem goes away. Then maybe you will need to connect the grounds of all the capacitors to a separate ground path by cuts and jumpers on the PCB.

I work with very high current equipment, and it can be surprising to read a significant voltage between two points on a piece of copper bus bar that is only a few microhms.

Good luck,

Paul

RL,

this is good, I'll get on this later tonight. A couple questions:

Can I just use old rca plugs with the signal and ground shorted or do I need a resistor in there?

On an adjacent shelf to the stereo gear.

Do you mean "with a ground wire from the isolation transformer casing or third wire socket pin connected." ? If so, and using it's chassis, I should just be able to use one of the exterior screws holding the chassis together?

I'll get on this as soon as time allows.

Thanks,

gary

Hi Paul,

One of the steps that contributed to our victory over the 120 Hz component of this hum problem was scratching out 3 sections of the ground tracks of the PCB around the filter caps. I have previously asked the question (which remains unanswered) why there are 2 ground tracks on the PCB, top and bottom, that are then joined back together at every solder point and standoff connection. I can see that the mounting holes are tinned from one side through to the other. Why not then just make the board with one ground track? What could I be missing here? Anyway, it wouldn't take much effort to separate the two circuits, I might need some advice on how and where to start and stop the two busses. I'll put together a diagram of the PCB with index numbers near the components, that way one could say "cut at 3 and 11, jumper that to 15..." and so on.

thanks,

gary

A short will function. If you have some biult with resistors less than

50K, they'll probably work here. Commercial terminators will likely be 50, 75 or 600 ohms.

A purchased isolation transformer will bring the ground wire through unbroken, with contact to the transformer body and housing. A connection to a screw terminal should work. If in Doubt check with an ohmmeter.

Notify me at the from address, with nospam removed, any time. The thread is getting buried.

RL

Mostly I suspect the capacitor labeled C106 on the schematic, which appears to be actually C304 and C305 on the PCB layout. There's also a C306 in the schematic that does not seem to be on the PCB. But it looks like the filaments are tied to a ground strip that is also used as the reference ground for the input and the output, and there will be current flow there that will generate a voltage and create hum.

The tracks to the filamants should be separated from the other ground tracks, and tied back directly to the negative side of the big filter caps, along with the anodes of D103 and D104. There should be nothing else connected to this separate ground. However, you might want to connect something like a 1k resistor from this filament return circuit to ground somewhere.

It's very confusing to have multiple parts with the same ref des. Usually for circuits with left and right channels, you mi8ght have R101, R102, C101, etc for the left, and R201, R202, C201, etc for the right. Components that are shared, like the power supply, could be labeled R1, R2, C1, etc, or R301, R302, C301, etc. Unfortunately the tubes are shared as A and B for left and right channel, so they seem to be labeled V301 and V302. It may have been better to have used one tube for the left and the other for the right. Actually i still don't see the reason for V301A.

I think just a resistor would do as well, if not better. If tubes act as transconductance, with output current proportional to grid voltage, then a resistor would provide a linear voltage. If V301A acts as a constant current source, I think you might get nonlinearity. But I have not really analyzed the circuit completely. By now I would have just replaced the whole thing with a wall-wart DC supply and a couple of ICs.

Paul

Thanks Paul,

I suspect that this preamp saw several evolutionary iterations before it reached me in kit form. I have also wondered about C306, which doesn't exist. Neither does C106. If you look on the line section schematic you will see an H+ schematic on the bottom that shows C304 and C305. As you have noticed some of this doesn't reflect what I was given to assemble, and the picture of the advertised preamp displays quite a few different components. For instance, I recently replaced the axial C301s and 303s with a radial version. It was suggested that the axial ones having a coil construction would be inducing current...needless to say I have replaced much of this thing already, including the tubes.

I made a little diagram of how I interpret your proposed changes here:

Is this correct?

gary

Well, it's one step in the right direction, but now there is no current path for the filaments back to the negative supply. So, remove the tracks on V300 and V301, pin 5, and run wires from these terminals back to the negative side of the big capacitors C304 and C305. The other grounds are probably OK.

What's with the crazy tube pin numbers? Each tube should be numbered 1 through 9.

See my other more recent post on an LT Spice simulation of this circuit using JFETs. For all the trouble with this circuit, you could build a decent preamp with about $5 worth of parts and have it fit in a matchbox. Why tubes? And if you want to be a tube purist, you really need to get rid of the silicon diodes :)

Paul

sorry about that, as I was starting to understand more about this stuff I found it easier to number the tube pins according to channel. I have since learned that this is not common practice but haven't corrected the drawing yet.

I'll check that out. I'd like to build this $5 preamp and compare. The honest truth is that I love music, I build my own speakers and I can't afford high end gear so I thought I could learn to build it. This has proven to be naive to say the least. Still, I am determined to go until I either give up or lose interest in listening to great sounding music. This preamp was an experiment, I wanted to know what the fuss over tube gear is all about. I thought that I might build an all tube system over a few years, some effecient speakers and give it a try. Those plans have changed a bit, I'm thinking of building the Linkwitz Orions next (hoping dipoles will get my neighbors off my back a bit,) so I'll first be building 8 50W chip amps to power each driver. I've seen several designs using the LM3875 and LM3886 chips. Do you have any experience with these amps?

gary

sorry about that, as I was starting to understand more about this stuff I found it easier to number the tube pins according to channel. I have since learned that this is not common practice but haven't corrected the drawing yet.

I'll check that out. I'd like to build this $5 preamp and compare. The honest truth is that I love music, I build my own speakers and I can't afford high end gear so I thought I could learn to build it. This has proven to be naive to say the least. Still, I am determined to go until I either give up or lose interest in listening to great sounding music. This preamp was an experiment, I wanted to know what the fuss over tube gear is all about. I thought that I might build an all tube system over a few years, some effecient speakers and give it a try. Those plans have changed a bit, I'm thinking of building the Linkwitz Orions next (hoping dipoles will get my neighbors off my back a bit,) so I'll first be building 8 50W chip amps to power each driver. I've seen several designs using the LM3875 and LM3886 chips. Do you have any experience with these amps?

I found some interesting links with a dogpile search for tube preamp:

$preamplifier.html

The more reasonable sites (particularly sweetwater) contend that the appeal of vacuum tube circuits is mostly the "pleasant sounding" even harmonic distortion that gradually appears when a tube circuit is overloaded, rather than the abrupt and harsh odd harmonic square wave distortion that occurs with solid state circuits.

More of the vacuum tube cultism is evident in the $99 preamp site, and the bottlehead site, although I must admit that there is a serious "cool" factor with the finely handcrafted enclosures and exotic glass tubes that glow and emit comforting infrared rays. But the description of the sound "qualities" seems more like wine tasting, where minuscule impurities are actually celebrated, and the ability to detect such nuances is the big deal.

For myself, I have some hearing loss, so extreme hi-fi is wasted on me. I have seen this thread as a learning experience, and it was interesting to refresh my knowledge of tube circuits. I also enjoyed translating the circuit to JFETs and running a simulation. If I had a real need for a preamp, I would probably put together the simple circuit and use it. But I have a lot of perfectly good audio gear that I have never (or very rarely) used, and I have other projects that demand my time.

I don't have any experience with the LM3875, but I see it has made it into the Wiki. It is a linear amplifier, which is inherently inefficient, and I have more interest in PWM type "Class D" amplifiers. See

I designed a 35 watt amplifier for a project some years ago, using push-pull darlington output transistors and an op-amp. It was for 50/60 Hz only, but it had to produce 120 VAC as well as 24 VAC, with short circuit protection. For another application, we used a 100 watt monolithic power Op-Amp, which had to operate up to 500 Hz. These were linear amplifiers, and I was working on various ways to increase efficiency. My early efforts to design a PWM amplifier (in the early 1980s) was unsuccessful, but I learned a lot.

For raw power, I remember seeing an Unholtz-Dickie shaker amplifier at the place where my father worked in the mid 1960s, and it had vacuum tubes that were probably 3 feet high and a foot in diameter. This was for testing components of missiles for survival under extremes of mechanical vibration and shock. The amplifier was rated at several kW, and was water-cooled. I thought it was a shame to use boring sine waves of fixed frequencies. I said they should just play some rock and roll! The company still exists, and they have a 480 kW shaker. Now that's some amplifier!

Let me know how your project turns out. I think many of us learned something from this discussion.

Paul